Replies: 1 comment
-
facing same type of issue while training the paddle detection model with custom dataset |
Beta Was this translation helpful? Give feedback.
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
Hi everyone,
I've been working on training PaddleOCR KIE on a custom dataset consisting of ID Cards, but I'm encountering a problem during testing. The training process completes successfully without any errors, but when I evaluate the model using sample images, it provides incorrect results.
Here are some details about my setup:
Despite these efforts, the model fails to produce accurate results during testing. I suspect that there might be something wrong with my configuration, training parameters, or data preprocessing steps.
I would appreciate any guidance or suggestions from the community on the following points:
Are there any common pitfalls or known issues when training PaddleOCR KIE on custom datasets?
What are the recommended strategies for debugging such issues?
Are there any specific hyperparameters or settings that should be adjusted for ID Card recognition?
Could there be any data preprocessing steps that I might be missing, causing the model's poor performance?
I'm open to any advice, insights, or pointers to relevant resources that could help me resolve this problem. Thank you in advance for your assistance!
Best regards,
Merkelon
Beta Was this translation helpful? Give feedback.
All reactions